This paper explores the dispute between the Republic of Azerbaijan and the Republic of Upper Karabakh with regard to the Sarsang reservoir and the associated water management by the latter. After illustrating the politicization of the environmental challenge by recent hostile initiatives taken by the parties, the article points out the lack of reliable statistics on water usage and needs for riparian countries. It highlights the matter of associated semantics as well as the poor commitments of these countries with regard to international water conventions. Finally, the paper suggests that the water issue could be disconnected from the political conflict and discusses the possibility to capitalize on past similar transboundary water issues to overcome the obstacles. It concludes on a possible process for relevant international technical committees to embark upon a separate water-related mediation based on water needs as a ‘public good’.

Geographical context and political outline

Sarsang reservoir is an artificial water basin located in the Republic of Upper Karabakh (40.2 N, 46.6 E). It is built on the Tartar River and is 12 km long and 0.6 km wide (see Figure 1). Its capacity is 575 million cubic meters (National Statistical Service of the Nagorno Karabakh Republic, 2010). It was formed in 1976, i.e. during the Soviet time, after the construction of a 125 m-tall hydroelectric dam.
Fig. 1.

A recent view of the Sarsang reservoir.

Fig. 1.

A recent view of the Sarsang reservoir.

Upper Karabakh is presently a self-determined State, east of Armenia and west of Azerbaijan, in the South Caucasus (see Figure 2). This mountainous area was continuously populated by ethnic Armenians. It was part of the 1st Republic of Armenia (1918–1920). However in 1921, when the Soviets annexed Armenia along with the 1st Republics of Azerbaijan and Georgia, they detached Karabakh from Armenia to attach it to Azerbaijan in order to please Turkey, which was then being tempted to join the Soviet Union. This was enacted in 1923 by the creation of the ‘Nagorno-Karabakh Autonomous Oblast’ (Upper Karabakh Autonomous Region)1, and entrusted the administration of the region to Azerbaijan Soviet Socialist Republic. This administrative attachment has never been really accepted by the inhabitants of the Region. As a matter of fact, when the USSR started collapsing in 1989, the Supreme Soviet of Upper Karabakh voted for the independence of their region2. This vote was confirmed in 1991 by the population of the Region through a referendum at the end of the same year, just before the dissolution of the USSR, and in accordance with the provisions of the then-enforced Soviet Constitution.
Fig. 2.

Upper Karabakh is landlocked between Armenia and Azerbaijan.

Fig. 2.

Upper Karabakh is landlocked between Armenia and Azerbaijan.

The fierce opposition of Azerbaijani authorities and the ensuing pogroms against Armenians living in Azerbaijan – in Karabakh and elsewhere – triggered an open war between the self-determined Republic and Azerbaijan forces. In this conflict, Armenia was then and is still allied with Karabakh. The strong resistance of Karabakh, internal intrigues in Azerbaijan and the alliance shift of Russia fostered the defeat of the Azerbaijan army. Azerbaijan authorities asked for a ceasefire, which was signed on 12 May 1994 by the three fighting sides, namely the self-determined Republic of Upper Karabakh, the Republic of Armenia, and the Republic of Azerbaijan.

The ceasefire-frozen frontline encompasses the territory of Upper Karabakh plus a security buffer zone, and minus certain territories of the former Soviet ‘Nagorno-Karabakh Autonomous Oblast’ claimed by Upper Karabakh and which remain under Azerbaijani control. Since 1994, Armenia and Azerbaijan have been negotiating a lasting peace with the help of the ‘Minsk Group’ (chaired by France, Russia and the United States) under the auspices of the Organization for Security and Cooperation in Europe (OSCE). Several summits (Key West 2001, Madrid 2004, Helsinki 2008, L'Aquila 2009) helped to define and consolidate the ‘Madrid principles’ that are supposed to outline the negotiations. Strong oppositions persist, however, between the parties about the way to implement these principles and, most importantly, about their order of implementation (phased or package settlement).

Currently, the process seems to be in a deadlock. More and more often the ceasefire is violated and shooting incidents occur practically on a daily basis between opposing armed forces, leading to casualties and deaths. Some observers consider that no peace settlement will be reached without the explicit participation of the self-determined Republic of Upper Karabakh in the negotiation, a perspective that is strongly opposed by Azerbaijan.

More detailed recalls on this conflict and on its historical background would be far beyond the scope of the present article. Interested readers may find detailed information in the literature on the regional history from ancient times, on the conflict and its genesis since the 19th century and after the Soviet rule over the region, and on the current negotiation process toward a lasting settlement (for instance, Kambeck & Ghazaryan (2013); De Waal (2013); Geukjian (2014)).

With regard to the water issue, it is noteworthy that the conflict by itself did not lead to some major ecological disaster or irreversible damage. But, from the prospect of natural resources, the new political situation had some consequences on the water supply of impacted countries, detrimental to Azerbaijan and beneficial to Karabakh. Because of its elevation, the wetlands of Karabakh were used as a kind of natural water reserve by the Azerbaijani authorities. This was especially true for the Tartar River and the Sarsang reservoir (see Sarsang location in Figure 3). For instance, an Azerbaijani author argued that
Fig. 3.

Location of reservoir in the northern part of Upper Karabakh. Map adapted from Peter Fitzgerald (2010), available at: https://commons.wikimedia.org/wiki/File:Nagorno-Karabakh_regions_map.png.

Fig. 3.

Location of reservoir in the northern part of Upper Karabakh. Map adapted from Peter Fitzgerald (2010), available at: https://commons.wikimedia.org/wiki/File:Nagorno-Karabakh_regions_map.png.

due to occupation of Sarsang reservoir with 560 million m3 capacity by the Armenian Republic, irrigation water supply became impossible on the 100,000 ha of agricultural land that caused irrecoverable losses to the economy of this region’ (Imanov, 2007).

Conversely, officials from the Upper Karabakh authorities claimed that

when Artsakh [Upper Karabakh] was still an autonomous region, Azerbaijan deprived it of [the right] to use the water resources. It built, for example, dams on the rivers of the former Nagorno-Karabakh Autonomous Region, but they were located beyond our borders. Hence, we weren't able to use those rivers’ (Tert.am, 2014b).

In the context of such global hostility, politicizing the allocation of water from the Sarsang reservoir became unavoidable. Analyzing this politicization and how to overcome it is the main point of this article.

Framework and objectives

The present work is a case study. It points out how a water resource issue has been politicized in the context of a wider regional conflict. It also discusses the various factors that hinder any clear assessment of the true needs and wills of the conflicting parties and, consequently, any step toward an agreement on water usage. In this regard, the Sarsang issue challenges the theoretical rationale of ‘water wars’ (Klare, 2002) or, conversely, ‘water rationality’ (Alam, 2002). This conflict does not fall easily into these categories for a wide set of reasons.

First, there is no severe water scarcity in the region, neither an overexploitation of the water resources due to increase of demographic pressure nor strong industrialization build-up. Though limited to Azerbaijan, Verdiyev (2004) gives, for instance, a sketch of the diversity and abundance of water resources in the region. Therefore, the water wars rationale based primarily on water scarcity is hardly justified – though other factors such as a wider conflict and bellicose public statements actually prevail. One can thus expect that water rationality based on a mutually beneficial cooperation could dominate, despite unresolved issues and surrounding disputes.

However, such rational behavior is not presently being explored. Actually, the political context and the disposition of the conflicting parties make it difficult to incite them toward such rationality. It has been underlined that rationality usually stems from both expected financial benefits – for instance, from the international community – and from a true interest in securing long-term water access for riparian countries (Alam, 2002).

In our case, both points are nullified. Azerbaijan is an oil-rich country that is not considered by its own authorities as needing any external financial support whereas Upper Karabakh, as a non-recognized State, is considered as not eligible for support from dedicated international institutions. Moreover, as water resources are not so scarce in the regions, access to Sarsang is not as critical as it may have been for the Israel–Jordan or the India–Pakistan case studies exemplifying the water rationality. Therefore, water is not and has never been the primary objective of the conflict, and related infrastructures have not been especially damaged deliberately or even accidentally. As a consequence, the Sarsang case is not really a water conflict where related analytical tools and framework may apply but rather a situation where water is more a ‘political tool’, i.e. where ‘water resources or water systems themselves are used by a nation, state or non-state actor for a political goal’ (Gleick, 2013).

This article primarily documents data and statements testifying the politicization process of this case that is mostly not referenced in the literature. Section 3 compiles the Sarsang-related political statements and initiatives giving account for the enduring bellicosity between the parties. It aims at providing evidence about the politicization of this specific issue, notwithstanding the wider hostility between Azerbaijan and the Upper Karabakh.

Beside this documentation and the probable lack of political will it reveals, the article explores the secondary obstacles that could hinder any progress even in the case of political improvement. Without exhausting all the obstacles, one can address three main categories of problems:

  • A knowledge issue: the lack of objective data on water usage by both riparian countries.

  • A semantic issue: linkage to internal politics leading to consideration of ‘rights’ rather than ‘needs’.

  • A legal issue: the poor commitment of the parties to relevant international conventions.

Sections 4, 5 and 6 address and illustrate these respective points, suggesting some possible ways to overcome the associated limitations in order to embark upon a mediation partly based on water needs as a ‘public good’.

Statements and initiatives toward politicization

Though the territorial dispute between Azerbaijan and the self-determined Republic of Upper Karabakh is not so recent, the Sarsang reservoir has gained focus and much attention only recently to become the object of many mutual claims and recriminations. Political communication strategies were developed in this regard.

On the Azerbaijani side, Mr Elkhan Suleymanov is one of the main actors of such strategies. He is a Member of Milli Mejlis (Azerbaijan Parliament), representing his country in the Euronest Parliamentary Assembly as well as in the Parliamentary Assembly of the Council of Europe (PACE) (Suleymanov, 2015). He is also the Chairman of the Association for Civil Society Development in Azerbaijan (ACSDA or AVCIYA in Azeri language), which appears to have been reactivated for the purpose of Sarsang in May 2013. Although it previously existed (Silverstein, 2007), ACSDA was rebranded as stemming from the civil society. One of its first initiatives was to send letters of complaint from concerned Azerbaijani citizens to the European Parliament3 (Azerbaijan Monitor, 2013a). Then, on 7 September 2013, ACSDA held a conference in the Tartar district (Azerbaijan) to raise the issue of the reservoir and the multidimensional threats that it put on the inhabitants of the neighboring districts of Azerbaijan (Azerbaijan Monitor, 2013b). The threats listed by ACSDA were the following:

  1. The reservoir and the dam have lacked any maintenance for 20 years.

  2. The dam is therefore likely to collapse, due to natural causes, technical and/or deliberate actions.

  3. The downstream region would experience a major environmental catastrophe, with the flooding of a population estimated to be 400,000.

  4. Karabakh authorities would have opened the reservoir in the winter, and hence deliberately provoked land flooding and road erosion.

  5. They would have also deliberately closed the dam in the summer thus drying agricultural lands and triggering serious ecological, agricultural and eventually social and economic tensions.

In the momentum of the conference, ACSDA and Mr Suleymanov launched a political initiative in the framework of the PACE. As early as August 2013, press releases claimed that a resolution was being prepared and even ‘approved’ by a PACE committee (Azeri-Press Agency, 2013a, 2013b, 2013c). This information was acquired by the Armenian press, which was later denied by a disclaimer from the PACE (Armenpress, 2013).

Actually, a working document was prepared by Mr Suleymanov. It was signed by roughly 50 members of the PACE. As this quorum was sufficient (PACE, 2015), the working document became a motion for a resolution – document 13270 (PACE, 2013) – to be handled by a PACE committee, among several others prepared by the ACSDA chairman in the same spirit (Suleymanov, 2015).

Interestingly, this initiative triggered a proposal from the official of the self-determined Republic of Upper Karabakh (Lragir, 2013). Arthur Aghabegyan, the Deputy Prime Minister suggested that his country and Azerbaijan could engage in a joint management of the Tartar River and of the Sarsang dam. Aghabegyan stressed that ‘the Sarsang reservoir has more capacity than is currently used. Both the Armenian and Azerbaijani sides will gain from proper use of the water canals that were built during the Soviet era’ and therefore ‘if Azerbaijan reacts positively to the proposed cooperation, the Armenian side will be responsible for financing large investments in the territories under its control in order to reconstruct the water resources’ (Aslanov, 2013). Apparently this proposal was inspired by the Georgian example: some weeks before, ‘the Georgian prime minister's economic advisor Georgy Khukhashvili stated that the Abkhazian railway can be operated without finalizing the status of Abkhazia’ (Lragir, 2013).

Azerbaijan did not reply to the position expressed by Aghabegyan and shared by David Babayan (Aslanov, 2013), a close adviser of President Bako Sahakian. Actually, Arif Yunus4, the prominent Head of the Department for Conflict Studies and Migration from the Institute of Peace and Democracy of Baku warned that ‘it is naive to think that Azerbaijan will accept that suggestion in a situation when the conflicting parties care for the political side of the issue, first and foremost’ (Arminfo, 2013).

As the counter-proposal made by Karabakh authorities seemed to undermine the position of Azerbaijan, nothing occurred for six months when Baku came back with the very same Azerbaijani position as before, through the same ACSDA. This time ACSDA set up a comprehensive website (www.sarsang.org) that gave exhaustive details on its positions, including the letters sent by citizens of Azerbaijan to either the European Parliament or to the PACE.

In this new initiative, three points are noteworthy:

  • A new sixth recrimination appears on the list, i.e. that ‘water [flushed from the reservoir] often washes away the landmines buried in the area of deployment of troops’ (News.az, 2014).

  • The Azerbaijani side attempted to mix the Sarsang issue with a wider set of problems. A direct example of such an attempt is given by a press release that intertwines Sarsang with the Nakhichevan issue, a district of Azerbaijan that is not contiguous to the Republic of Upper Karabakh and which is a territory not disputed (Azeri-Press Agency, 2014).

  • The overall result of this activism led to a new motion for a resolution (PACE, 2014a), which shifts the focus to ‘Inhabitants of frontier regions of Azerbaijan are deliberately deprived of water’.

This second motion for a resolution received greater attention from the relevant PACE Committee on Social Affairs, Health and Sustainable Development since this committee appointed a rapporteur whereas previously it just took note of document 13270. On 12 May 2014, Mrs Milica Marković (Bosnia and Herzegovina, Socialist) was appointed to report until April 2016 (PACE, 2014b).

Almost at the same time, the proposal of the Upper Karabakh side for a joint management received an unexpected support. After a short visit to Karabakh, James Warlick, the US co-chairman of the OSCE Minsk Group, posted on Twitter that ‘it would be a positive step if the sides could jointly manage water resources’ (Warlick, 2014), an assessment that immediately received bitter comments from various Azerbaijan officials. In particular, Mr Suleymanov termed this proposal as ‘fantastic and crazy ideas’ (Azerbaijan Monitor, 2014).

It seems that the initiative of Ambassador Warlick was not only tactical but actually represented a move of the Minsk Group toward a step-by-step resolution of the conflict. A few days before, Ambassador Warlick expressed his concern that ‘the status quo is unsustainable and increasingly dangerous’ and he could have thought that such a move regarding Sarsang could defuse the tension. Anyway, this proposal quickly became an official position of the OSCE Minsk Group when the three co-chairmen ‘visited the Sarsang reservoir, and discussed its status and operations with managers of the facility’ to express ‘their hope that the sides will reach an agreement to jointly manage these water resources to the benefit of the region’ (OSCE Minsk Group, 2014).

This new position from Ambassador Warlick and the OSCE Minsk Group was positively assessed by the Upper Karabakh authorities and revived the proposal they made one year before (Tert.am, 2014a). In a dedicated interview, David Babayan gave some clue on the way he is considering such a possible cooperation. He stated that this offer was ‘of course, a manifestation of good will’ but with ‘an economic component’. He emphasized that Karabakh is

not a charity-maker to gratuitously offer our product to a state which treats us in a hostile way, to say the least. But we understand we are neighbors, so we have to co-exist as independent states. That is why we say, ‘let us establish a mutually beneficial cooperation for the beginning’. I repeat again that this cannot be a joint use; just a mutually beneficial cooperation.’

He also mentioned some precedents of such cooperation between conflicting countries, citing the ‘Peace channel’ for the Dead Sea between Israel, Jordan and the Palestinian authority, though recognizing that ‘there are problems there too’. He concluded by affirming that the proposal for cooperation was part of official Karabakh proposals since 1998–1999 and that this policy was contrasting with Azerbaijan's ‘very tough hydrogeopolitics policies’ that deprived his country of water resources. Babayan (Tert.am, 2014b) explained that, during the Soviet time,

the Sarsang waters were used to irrigate lands with an area of 128,000 hectares, of which only 10,000–15,000 hectares were for Artsakh (i.e. Upper Karabakh). So the rest went to Azerbaijan which controlled the management network.’

Asked whether Nagorno-Karabakh makes a full use of the reservoir's water resources, Babayan answered that they have such a chance only during the winter months. He explained,

Yes, we make a full use of it for the energy production, but we are forced to use it in winter, as the water also goes to Azerbaijan while the power is being generated. So we use in winter, and they cannot use it in summer.’

Babayan's affirmations seem to be confirmed at a technical level as Yuri Javadyan, director of ArmWaterProject Company, presented a project of Sarsang Water Reservoir's water usage for irrigation to Ara Harutyunyan, Prime Minister of Upper Karabakh. According to Javadyan (Arka.am, 2014), ‘18,000 hectares of agriculturally used land in Martakert province will be irrigated by water from this reservoir’.

Nevertheless, Babayan in his conclusion leveraged the politicization allowed of the water resource and insinuated that Azerbaijan should come quickly to an agreement. He stated that

Sarsang has to be used for hydropower in the first place and irrigation – in the second. We cannot use it for irrigation yet, as the networks have remained there. Activities are now underway for restoring the system completely. So, Azerbaijan still has time to agree, because after we restore the water distribution networks, we will no longer need an economic cooperation with Azerbaijan’ (Tert.am, 2014b).

Lack of data on Sarsang water usage

There are numerous examples of conflicts related to water supplies. Obviously, they are not all of the same nature. Some of them are linked to natural causes such as water scarcity, droughts and climate changes. Some others are due to human interventions such as dams or transboundary usage mismatch. Databases (Oregon State University, 2015) and analysis – as illustrated by studies often granted by US or EU agencies (Wolf et al., 2003; Yoffe et al., 2003; Gleditsch et al., 2006) – attempt to define conflict typologies. In summary, disputes or conflicts arise when different human groups and their authorities – whether recognized States or not – disagree on the way to manage a given water basin. In this regard, it has often been underlined that water dispute is an alibi for larger and deeper disagreements or hostility, or as a negotiable issue in interstate trade-offs.

It has also been underlined that focusing on a very local water issue is probably not a good way to deal with the global problem. It is more appropriate to consider the relevant hydrographic unit, i.e. the water basin. In our case, the Kura-Araks basin has globally been identified as being ‘at risk’ (Wolf et al., 2003). Since the fall of the USSR, the Kura-Araks basin has been shared not only by Armenia, Azerbaijan, and Georgia but also by self-determined entities such as Upper Karabakh and partially South Ossetia. Turkey and Iran also share part of the basin though they are less concerned. Actually Upper Karabakh is at the very center of the basin, and is the eastern part of the Armenian highlands overtopping the Azerbaijan plain eastward (see Figure 4).
Fig. 4.

The Kura-Araks basin according to Vener (2006). Upper Karabakh area added by the author of the present article.

Fig. 4.

The Kura-Araks basin according to Vener (2006). Upper Karabakh area added by the author of the present article.

In the past, several surveys have been carried out on the Kura-Araks basin but they are now hardly exploitable: on the one hand, data are old – if not scarce – and would need to be revisited. On the other hand, due to the lack of international recognition, Upper Karabakh is ignored by these surveys (as exemplified by Table 1) so that its shares in the water inflows, outflows, needs and usages are largely not compiled.

Table 1.

Existing hydropower stations in Azerbaijan according to United Nations Development Programme (2013). Sarsang is still considered among hydropower stations existing in Azerbaijan.

HPP River Capacity (MW) Annual energy production (GWh) Water discharge (m3/s) 
Mingechevir Kura 370 1,355 780 
Shamkir Kura 380 845 850 
Varvara Kura 17 90 360 
Yenikend Kura 150 395 810 
Aras Aras 260 80 260 
Sarsang Terter 50 123 67 
Vaykhir Nakhchivanchay 12 11 
Total  1,232 2,900  
HPP River Capacity (MW) Annual energy production (GWh) Water discharge (m3/s) 
Mingechevir Kura 370 1,355 780 
Shamkir Kura 380 845 850 
Varvara Kura 17 90 360 
Yenikend Kura 150 395 810 
Aras Aras 260 80 260 
Sarsang Terter 50 123 67 
Vaykhir Nakhchivanchay 12 11 
Total  1,232 2,900  

For instance, in a pioneering work focused on ‘Water Quality and Public Health’, Ewing just noted more than a decade ago that

in Armenia, the primary use of surface water from the Araks River is for agriculture and industry, while farther downstream in Azerbaijan, the Kura-Araks River is relied upon for drinking water as well as for agriculture and industry […] Azerbaijan is short on water, only allowing an average use of 1000 m3 per person per year, which is one of the lowest rankings in the world […] The shortage of water resources in Azerbaijan is compounded by their inefficient use. Broken-down irrigation systems lead to water losses of up to 50%’ (Ewing, 2003).

These observations were basically not revisited by Vener (2006) who endorsed nearly the same assessment.

Even in 2012, revisited data were non-existent. For instance, Campana et al. were unable to account of such data though they clearly gave account of substantial changes such as the fact that

each country has adopted a water code since 1992: Armenia in 1992 and revised in 2002 according to the European Union Water Framework Directives; and Georgia and Azerbaijan in 1997. Nevertheless, there is no uniform control and/or management system for the rivers and, in the post-Soviet period, no water quality monitoring by the riparian countries since 2002’ (Campana et al., 2013).

And if new data were compiled (United Nations Development Programme (UNDP), 2013; Taslakyan, 2014), they also failed to take account of data specific to the Upper Karabakh. Indeed, such a first attempt was made to assess the true situation on the ground by satellite imagery (Baumann et al., 2015). However, it was on war-induced land-use modifications and not on water usage and needs.

A serious step forward would thus be to reconsider the true rationale for water usage by analyzing related data for all States or States-like entities. In this regard, Babayan's expressed positions (Tert.am, 2014b) should be investigated and substantiated by statistical data on water usage both in summer and winter. Any possible future agreement on water management and exchange would thus be based on fair assessment of water needs and production. A first attempt in this regard has recently been offered by Dietzen (2014) who provides new data and some recent statistics on water usage by the Republic of Upper Karabakh. He also clearly exposes how Shahumian and Kashatagh districts of Upper Karabakh are essential for water security of both this country and neighboring Armenia and how it could be a tool for cooperation with Azerbaijan in the context of a ‘water for oil’ or ‘water for gas’ process.

However, embarking upon this kind of rationale means endorsing the assumption that water is a commodity that can be traded in the same way as oil or other primary resources. This conception is quite usual but it may also be challenged by non-merchant views. In this regard, words through which the issue is dealt with appear to be critical.

From water rights to water needs, from rhetoric to resolution

The most well-known water disputes are those where open or creeping conflicts are pre-existing and where water is actually a scarce resource, for instance, the Jordan River and Dead Sea basin, Mesopotamia with Tigris and Euphrates crossing Turkey, Syria and Iraq, Central Asia with the Aral Sea and the related Amu Daria and Syr Daria rivers. Detailed analyses of those water conflicts have already been extensively presented (Giordano et al., 2002).

However, some clear learning may be drawn from these cases and from the related confidence building measures (CBM) that were initiated. Perhaps one of the most important lessons to learn is that any attempt to tightly bind a given environmental concern to a political issue is the surest way to derail any CBM. Addressing the Tigris–Euphrates rivers basin issue between Turks, Kurds and Syrians (prior to the rise of the Islamic State), Voza et al. (2012) stated that ‘the separation of water issue from the wider security concerns […] and a focus on region under-development and optimal utilization of water resources, could be an instrument in building regional cooperation in the Middle-East’. Though the authors immediately add that ‘unfortunately, the prevailing wisdom suggests that the solution of an international environmental problem rarely comes before resolving wider political conflicts’ (Voza et al., 2012).

Obviously, in our case, one can tautologically assume that such a conflict directly stems from the previous situation of war between Azerbaijan and Upper Karabakh. However, as recalled by Voza et al. (2012), water conflicts may also suffer from linkage to ‘domestic political goals’. Actually, the open war in the South Caucasus ended a number of years ago but the Sarsang issue seems to be artificially perpetuated by the Azerbaijani regime for the internal purpose of diverting the attention of its population. As the regime is regularly noted among the most repressive in the world (see, for instance, World Press Freedom Index (2014)), a conciliatory way to embark upon a dialogue with Stepanakert can thus hardly emerge at a public level, even locally or for non-governmental organizations (NGOs).

This is particularly detrimental as over the last decade there has been a global trend to shift from ‘water rights’ to ‘water needs’, a perspective in which the nearby population of Azerbaijan could rightfully claim to benefit from Sarsang as was the case during the Soviet time. However, for Karabakh, such ‘needs’ are lacking reliable and recent data, because Azerbaijan cannot be deemed as democratic enough to issue data that are not politically biased (World Press Freedom Index, 2014; Human Rights Watch, 2015).

Indeed, such a possible bias in the claimed needs is not unusual: as recalled once again by Voza et al. for Turkey and Syria,

there is a great deal of controversy about way the ‘irrigable land’ figures are calculated and whether or not there has been strategic manipulation of these data […] Therefore, it will not be only uneconomical, but it will also be inequitable, to utilize scarce water resources to irrigate infertile lands at the expense of fertile lands’ (Voza et al., 2012).

Another way to reach a fair assessment of the needs would thus be to democratize Azerbaijan, a very political issue that goes far beyond a water issue.

This prerequisite could allow entering into a fair water allocation process. Such a process would nevertheless stem from the idea that water is a public good and not a commodity (Voza et al., 2012) and the Karabakh authorities could be reluctant to admit this conception. After all, one can wonder why water should be considered as a ‘public good’ but not energy, i.e. oil from which both Upper Karabakh and Armenia are largely deprived: as a post-war retaliatory measurement, Azerbaijan deliberately bypassed the two countries when it implemented its energy transport infrastructures.

For instance, considering a quite similar case in this regard, the opinion of the former Prime Minister of Turkey, Suleiman Demirel, was

neither Syria nor Iraq can lay claim to Turkey's rivers, any more than Ankara could claim their oil. We have a right to do anything we like. The water resources are Turkey's, the oil resources are theirs. We don't say we share the oil resources, and they cannot say they share our water resources’ (Alam, 2002).

Considering water and/or oil as a ‘public good’ or as a commodity thus underlines the importance of wording in this kind of conflict. This language issue has been extensively analyzed for the Israel–Jordan water settlement by Fischhendler et al. (2013) (see Table 2). In their conclusion, these authors underscored that

much of the deadlock was resolved only when the parties moved from their adversarial positions to address the interests behind the positions, where a compromise was forged that captures elements of international law while still addressing the needs of the stronger riparian – for example, adopting rightful allocation terminology in the case of Israel and Jordan, and rights based on needs in the case of Israel and the Palestinians. The ‘rights’ terminology came to satisfy the Jordanians or the Palestinians while the ‘allocation’ or the ‘needs’ terminology came to address the Israeli needs. The Red-Dead talks also exposed an integrative stage of negotiation during which the parties started to add benefits to the agreements. This is the ‘beneficiary party’ definition, which helped bypass any allocation and recognition based upon water ‘rights’’ (Fischhendler et al., 2013).

Table 2.

Evolution of negotiations over the Red Sea agreement (Fischhendler et al., 2013).

Content 2002 Draft 2003 Draft 2004 Draft 2005 Agreement 
Aim - Save the Dead Sea - Save the Dead Sea - Save the Dead Sea - Feasibility 
- Peace - Peace - Peace 
- Desalinization - Desalinization - Desalinization 
- Energy - Energy 
Scale Not defined Implicitly the whole basin Lower basin Water conveyance route 
Parties definition Governments Riparian; parties Beneficiary parties Beneficiary parties 
Parties involved Israel and Jordan All basin riparian Israel, Jordan, Palestinians Israel, Jordan, Palestinians 
Nature of cooperation Joint examination Regional institutions and possibly joint management To be studied To be studied 
Scope Not defined - Red–Dead conveyance and alternative routes - Red–Dead conveyance and alternative routes - Red–Dead conveyance 
- Water resources management - Water resources - Water policy statements 
Effect on water and land rights Implications on land and water rights Implications on land and water rights Not to ‘prejudice the riparian rights’ Not to ‘prejudice the riparian rights’ 
Public Participation Israel, Jordan, World Bank, USAID, US State Department Consolidation with a wider public Consolidation with a wider public Consolidation with a wider public 
Content 2002 Draft 2003 Draft 2004 Draft 2005 Agreement 
Aim - Save the Dead Sea - Save the Dead Sea - Save the Dead Sea - Feasibility 
- Peace - Peace - Peace 
- Desalinization - Desalinization - Desalinization 
- Energy - Energy 
Scale Not defined Implicitly the whole basin Lower basin Water conveyance route 
Parties definition Governments Riparian; parties Beneficiary parties Beneficiary parties 
Parties involved Israel and Jordan All basin riparian Israel, Jordan, Palestinians Israel, Jordan, Palestinians 
Nature of cooperation Joint examination Regional institutions and possibly joint management To be studied To be studied 
Scope Not defined - Red–Dead conveyance and alternative routes - Red–Dead conveyance and alternative routes - Red–Dead conveyance 
- Water resources management - Water resources - Water policy statements 
Effect on water and land rights Implications on land and water rights Implications on land and water rights Not to ‘prejudice the riparian rights’ Not to ‘prejudice the riparian rights’ 
Public Participation Israel, Jordan, World Bank, USAID, US State Department Consolidation with a wider public Consolidation with a wider public Consolidation with a wider public 

USAID: United States Agency for International Development.

Beyond any doubt, embarking upon any negotiation process on Sarsang water usage would necessarily imply a prior agreement on words and the associated legal framework. It would be a great step forward to succeed in solely implementing a joint Azerbaijan–Karabakh technical commission appointed to figure out a suitable wording and the legal framework before any direct negotiation. From a political standpoint, this commission could valuably be patronized or supported by the OSCE Minsk Group, in charge of supporting the conflict resolution.

Considering legal aspects

However, talking about a legal framework when it comes to Azerbaijan and Upper Karabakh is a challenge in itself. A quick look on the huge database of International Freshwater Treaties (Oregon State University, 2015) shows that only five treaties are directly related to the Kura-Araks basin. But they are mostly outdated and irrelevant as they deal with agreements between Iran and the Soviet Union on the Araks River. As recalled by Taslakyan (2014), the Soviet Union signed up some other water-related treaties with Turkey, and after the fall of the USSR, new agreements were signed between the various new independent States and their neighbors: between Azerbaijan and Georgia ‘in the field of the protection and sustainable development use of water resources of the Kura River Basin’, between Armenia and Turkey ‘on the use of Akhuryan water reservoir’, and between Armenia and Georgia ‘in the sphere of environmental protection and Natural Resources’. One can, however, notice that, regardless of Karabakh, there is not a similar agreement between Armenia and Azerbaijan.

Besides, the three internationally recognized countries of the South Caucasus have accessed the United Nations Economic Commission for Europe (UNECE) and, as such, they have signed or ratified some Multilateral Environmental Agreements (MEAs). In this regard, Armenia is less advanced than Azerbaijan as the former went through the first cycle of the UNECE's Environmental Performance Review (EPR) whereas the latter went through its second cycle. In this context, two important legal instruments must be considered: the Helsinki Convention on Protection and Use of Transboundary Watercourses and International Lakes (UNECE, 1992) and its additional Protocol on Water and Health (UNECE, 1999). In coherence with the EPR assessments, Armenia just signed the additional protocol (17 June 1999) whereas Azerbaijan accessed both the Convention (3 August 2000) and its additional protocol (9 January 2003). It is worth mentioning that the Convention itself was amended to allow the accession of non-UNECE member-States (UNECE, 2003). However, these amendments clearly limit themselves to UN-recognized States, thus excluding the Republic of Upper Karabakh. From a legal perspective, the situation seems thus to be relatively blocked: Armenia hasn't accessed the instruments yet and Upper Karabakh is not allowed to do so.

This is damaging, especially for Azerbaijan, as the Convention and its protocols enforce some rights for downstream riparians of transboundary watercourses. Of course, Upper Karabakh could unilaterally decide to comply with the provisions of these instruments but this process is unlikely to occur as it would not be balanced by any reciprocal move from Azerbaijan or from the international community. Especially, one point of concern for Karabakh is to gain recognition of the basic human rights of its own people, among which is the right to water. As previously recalled, during the Soviet period, Karabakh inhabitants were mostly deprived of the water flowing from their land for the benefit of the Azerbaijani plain. And even today, Azerbaijan doesn't seem to be ready to consider the right to water (or any other right) of the Karabakh people, regardless of the legal status of this entity.

Yet, in compliance with the idea that water would be a ‘public good’, there is a wide consensus on the fact that people may have a ‘human right to water’ (Office of the United Nations High Commissioner for Human Rights, 2010). This position stems from various UN conventions dating back to the 1970s (for a panorama, check the leaflet released by the United Nations (2014)) and it is now endorsed by most of the UN bodies and related international agencies such as the World Health Organization (WHO, 2003) or even the World Bank (Salman & McInerney-Lankford, 2004). However, there is a fundamental ambiguity as this right is supposed to be granted to every single human being whilst legally binding only State parties, and therefore only recognized States. The contradiction lies in the fact that recognized States are supposed to provide and enforce these rights whereas unrecognized bodies are often considered as potentially threatening them. For instance, a very detailed analysis on human rights obligations of non-state actors performed by Clapham (2006) is based on this assumption that ‘the threat to human rights posed by non-state actors is of increasing concern’. In our case, Karabakh authorities consider the situation the other way round and see Azerbaijan as a threat to human rights.

A possible way to start talking about the right to water – for Sarsang riparians of both sides – would be once again to disconnect talks on water from talks on the political status. Though his analysis goes far beyond our water-focused topic, and though it addresses more basic human rights, Clapham perfectly illustrates this position by explaining that

insurgents who were recognized by the State against which they were fighting not only as insurgents but also expressly as belligerents, became assimilated to a state actor with all the attendant rights and obligations which flow from the laws of international armed conflict […]’ (Clapham, 2006).

But he immediately adds that

today, international law imposes obligations on certain parties to an internal armed conflict irrespective of any recognition granted by the state they are fighting against or by any third state. The problem is that governments are often loath to admit that the conditions have been met for the application of this international law, for to admit such a situation is seen as an admission that the government has lost a degree of control and as an ‘elevation’ of the status of the rebels’ (Clapham, 2006).

He concludes quite optimistically that

political obstacles to a general shift in thinking are considerable: governments will accuse the humanitarians of bestowing legitimacy on their enemies and threaten to withdraw co-operation. But this challenge has been overcome with humanitarian law. No one should be accused of backing terrorists by accusing them of violating Common Article 3 to the Geneva Conventions. It is time to do the same with regard to human rights law. Once we rid ourselves of the assumption that human rights only cover the relationship between individuals and governments, there is no danger that accusing an armed group of human rights violations lends it automatic legitimacy or quasigovernmental status’ (Clapham, 2006).

Concluding remarks and recommendations

Recent developments show that the Sarsang reservoir and its water usage have become a source of bitter dispute between Azerbaijan and the self-determined Republic of Upper Karabakh. Baku is presently blaming this unrecognized State for its non-collaborative management of the water, i.e. for the very same reason that triggered the complaints of Karabakh people when their land was controlled by Azerbaijan. In this context of global hostility, it has led to a strong and seemingly counterproductive politicization of the dispute. Disconnecting the water management as a technical issue from the global issue of resolving the conflict would of course be a required step forward but it will probably not be sufficient.

First, there are no recent unbiased data on water usage, neither in Upper Karabakh, nor in Azerbaijan. For Upper Karabakh, such a lack of data may be linked to the fact that the State is both unrecognized and too flimsy: international organizations are thus ignoring it in water usage census campaigns and it is not able to perform them by itself. For Azerbaijan, the authoritarian nature of the regime casts strong doubt on the reliability and on the seriousness of data.

A possible approach for international organizations – for instance for UNECE – would thus be to negotiate with authorities of both countries – whether recognized or not – the way to achieve some independent and reliable water usage and water needs statistics in the various seasons. Such an initiative could be launched in full coordination with the OSCE's Minsk Group and it could even be strongly demanded by the chairmanship of this group, as recalled by Clapham, and it would not entail any political signification. In this regard, committees participating in UNECE – for instance, its two working groups, respectively ‘on Integrated Water Resources Management’ and on ‘Monitoring and Assessment’ – are probably the most qualified bodies for supervising this kind of work. In this context, the technical support could very valuably be brought by the ‘International Water Assessment Center’ (IWAC, 2015), which is closely associated with the UNECE water bodies.

Eventually, from this new data an agreement can be drafted on collaborative (if not ‘joint’) water management policies by both sides. Of course, one cannot ignore the major obstacles for such an attempt to build confidence. In this paper, I have non-exhaustively mentioned two obvious examples of such obstacles. On the one hand, past examples show how sensitive the wording issue could be and it would probably be a good idea to take inspiration from these previous examples, for instance, by associating actors that were involved from within to the Sarsang issue. On the other hand, the notion of water as ‘public good’ remains far from being commonly accepted, especially for Karabakh and Armenia. Being subjected to Azerbaijan-originated energy-depriving infrastructures, these two countries will probably be reluctant to grant the latter with water as a ‘public good’. In this regard, it is worth remembering Babayan's statement (Tert.am, 2014b), according to which Karabakh is ‘not a charity-maker to gratuitously offer our product to a state which treats us in a hostile way’. A more conventional negotiation of water against oil or gas is likely to emerge and one can actually wonder why water would be a ‘public good’ but not energy. A true question that goes far beyond the scope of this paper.

Acknowledgement

The author would like to thank Jirayr Momjian for a careful reading of the manuscript and for many insightful and valuable comments.

1

Beyond its 15 main Soviet Socialist Republics, the Soviet Union included a time-variable number of Autonomous Republics and Autonomous Regions (Oblast) – respectively 22 and eight in 1978. These autonomous republics and regions were granted differentiated levels of autonomy, though their administration remained assigned to one of the Republics.

2

The Soviet constitution allowed in theory that Republics and Regions could declare their exit from the Soviet Union but in practice this right to exit became possible only during Glasnost and Perestroika.

3

The Azerbaijan Monitor, source of this information, may be confusing the European Parliament with the Parliamentary Assembly of the Council of Europe (PACE) since the letters were apparently sent to the members of the PACE and not to those of the European Parliament.

4

Since then, Arif Yunus has been arrested with his wife Leyla during an unprecedented repression against human rights activists in Azerbaijan. See, for instance, Human Rights Watch website (Index of Censorship, 2014).

References

References
Arka.am
(
2014
).
Project of Sarsang Water Reservoir's water usage for irrigation presented to Karabakh premier, 1 September 2014. Available at: http://arka.am/en/news/economy/project_of_sarsang_water_reservoir_s_water_usage_for_irrigation_presented_to_karabakh_premier/ (accessed 12 November 2015)
.
Armenpress
(
2013
).
PACE denies Azerbaijani provocation regarding Sarsang Water Reservoir, 5 September 2013
.
Arminfo
(
2013
).
Arif Yunusov: It is naive thinking that Baku will accept Arthur Aghabekyan's suggestion
.
Aslanov
B.
(
2013
).
Azerbaijan applies for PACE Resolution on the Sarsang Reservoir in Nagorno-Karabakh. Central Asia – Caucasus Analyst
.
Azerbaijan Monitor
(2013a).
German Translation
, 29 June 2013.
Azerbaijan Monitor
(2013b).
German Translation
, 21 August 2013.
Azerbaijan Monitor
(
2014
).
American Envoy Accused of Siding with Armenia in Sarsang Water Dispute, 22 May 2014
.
Azeri-Press Agency
(
2013a
).
Motion on Sarsang reservoir, occupied by Armenia, to be reviewed at PACE Bureau meeting
, 28 August 2013.
Azeri-Press Agency
(
2013b
).
PACE Bureau approves motion on Sarsang reservoir in Azerbaijan's occupied territory
, 2 September 2013.
Azeri-Press Agency
(
2013c
).
PACE panel approves draft resolution on Sarsang water reservoir
, 12 September 2013.
Azeri-Press Agency
(
2014
).
Azerbaijan urges inspection of Nakhchivan's blockade, Sarsang humanitarian catastrophe
, 4 March 2014.
Baumann
M.
Radeloff
V. C.
Avedian
V.
Kuemmerle
T.
(
2015
).
Land-use change in the Caucasus during and after the Nagorno-Karabakh conflict
.
Regional Environmental Change
15
(
8
),
1703
1716
. Available at: http://link.springer.com/article/10.1007%2Fs10113-014-0728-3 (accessed 5 November 2015).
Campana
M. E.
Vener
B. B.
Lee
B. S.
(
2013
).
Hydrostrategy, hydropolitics, and security in the Kura-Araks Basin of the South Caucasus
.
Journal of Contemporary Water Research & Education
149
(
1
),
22
32
.
Clapham
A.
(
2006
).
Human rights obligations of non-state actors in conflict situations
.
International Review of the Red Cross
88
(
863
),
491
523
.
De Waal
T.
(
2013
).
Black Garden: Armenia and Azerbaijan through Peace and War
.
NYU Press
,
New York
.
Dietzen
M.
(
2014
).
Whose hands on the spigot? Water Security and the Nagorno-Karabakh Conflict
.
Streit Council for a Union of Democracies
,
Annandale, VA
,
USA
.
Ewing
A.
(
2003
).
Water Quality and Public Health Monitoring of Surface Waters in the Kura-Araks River Basin of Armenia, Azerbaijan, and Georgia, Water Resources Program
.
The University of New Mexico
,
WRP-8
, Albuquerque, New Mexico, USA. Available at: http://www.unm.edu/~wrp/wrp-8.pdf (accessed 5 November 2015).
Fischhendler
I.
Wolf
A. T.
Eckstein
G.
(
2013
).
The role of creative language in addressing political realities: Middle-Eastern water agreements
. In:
Shared Borders, Shared Waters
.
Megdal
S. B.
Varady
R. G.
Eden
S.
(eds).
CRC Press/Balkema
,
Leiden
.
Geukjian
O.
(
2014
).
Negotiating Armenian-Azerbaijani Peace: Opportunities, Obstacles, Prospects
.
Ashgate New edition
,
Farnham
,
UK
.
Gleditsch
N. P.
Furlong
K.
Hegre
H.
Lacina
B.
Owen
T.
(
2006
).
Conflicts over shared rivers: resource scarcity or fuzzy boundaries?
Political Geography
25
,
361
382
.
Gleick
P.
(
2013
).
Water Conflicts Database and Typology
.
The World's Water, Pacific Institute
,
Oakland, CA
.
Human Rights Watch
(
2015). Website. Available at: https://www.hrw.org/europe/central-asia/azerbaijan (accessed 5 November 2015)
.
Imanov
F. A.
(
2007
).
Water Infrastructure of Kura Araks Basin within Azerbaijan
.
International Congress on River Basin Management
, Antalya, Turkey.
Index of Censorship
(
2014
).
Letter: Persecution of human rights defender Leyla Yunus
.
IWAC-International Water Assessment Center
(
2015)
.
A Pan-European network of scientists and policy makers to support integrated water resources management
.
Kambeck
M.
Ghazaryan
S.
(
2013
).
Europe's Next Avoidable War: Nagorno-Karabakh
.
Palgrave Macmillan
,
Basingstoke, Hampshire
,
UK
.
Klare
M. T.
(
2002
).
Resource Wars: The new Landscape of Global Conflict
.
Henry Holt & Company
,
New York
.
Lragir
(
2013
).
NKR Official Calls For Cooperation With Azerbaijan
.
National Statistical Service of the Nagorno Karabakh Republic
(
2010
).
Nagorno-Karabakh in figures
.
News.az
(
2014
).
400,000 Azerbaijanis have become hostages of crash at Sarsang water reservoir
.
Office of the United Nations High Commissioner for Human Rights
(
2010
).
The Right to Water, Fact Sheet No. 35
.
Oregon State University, Program in Water Conflict Management, Transformation
( (accessed 5 November 2015).
OSCE Minsk Group
(
2014
).
Press Release by the Co-Chairs of the OSCE Minsk Group
.
PACE
(
2013
).
Expected humanitarian catastrophe caused by the dangerous condition of the Sarsang water reservoir in the occupied territories of Azerbaijan, Motion for a resolution Doc. 13270
.
PACE
(
2014a
).
Inhabitants of frontier regions of Azerbaijan are deliberately deprived of water, Motion for a resolution Doc. 13468
.
PACE
(
2014b
).
Committee on Social Affairs, Health and Sustainable Development, Reports under preparation AS/Inf 08
.
PACE
(
2015)
.
Rules of Procedure, Rule 24 – Tabling of motions for recommendations and resolutions
.
Salman
M. A. S.
McInerney-Lankford
S.
(
2004
).
The Human Right to Water, Legal and Policy Dimensions
.
The World Bank
, New York.
Silverstein
K.
(
2007
).
ACSDA seems to be previously identified as a Bogus Azeri ‘Human Rights Group’ dealing with ‘government propaganda mission’. Harper's Magazine
.
Suleymanov
E.
(2015). PACE website
. Available at: http://www.assembly.coe.int/nw/xml/AssemblyList/MP-Details-EN.asp?MemberID=6679 (accessed 12 November 2015).
Taslakyan
L.
(
2014
).
Transboundary Management of the Kura-Araks River Basin, Presentation made for the project on Reducing Transboundary Degradation in the Kura Ara(k)s River Basin, held by the Global Environment Facility (GEF) of the United Nations Development Programme (UNDP)
.
Tert.am
(
2014a
).
James Warlick: Joint use of water resources would be positive step
, 19 May 2014.
Tert.am
(
2014b
).
Nagorno-Karabakh's ‘water proposal’ to Azerbaijan's ‘tough geopolitics’
, 20 May 2014.
UNDP
(
2013
).
Updated Transboundary Diagnostic Analysis – Reducing Transboundary Degradation in the Kura Araks Basin
.
GEF Project Report
.
UNECE
(
1992
).
UNECE Convention on the Protection and Use of Transboundary Watercourses and International Lakes
.
UNECE
(
1999
).
UNECE Protocol on Water and Health to the 1992 Convention on the Protection and Use of Transboundary Watercourses and International Lakes
.
UNECE
(
2003
).
Amendment to Articles 25 and 26 of the Convention, Economic Commission for Europe Meeting of the Parties to the Convention on the Protection and Use of Transboundary Watercourses and International Lakes
.
United Nations
(
2014)
.
The Human Right to Water and Sanitation, Milestones, UN-Water Decade Programme on Advocacy and Communication
.
Vener
B. B.
(
2006
).
The Kura-Araks Basin: Obstacles and Common Objectives for an Integrated Water Resources Management Model among Armenia, Azerbaijan, and Georgia
.
University of New Mexico
, Albuquerque, New Mexico, USA.
Verdiyev
R.
(
2004
).
The role of ecosystems of Azerbaijan – forests and wetlands – as water suppliers
.
Seminar on the Role of Ecosystems as Water Suppliers
.
Voza
D.
Vukovic
M.
Carlson
L.
Djordjevic
D. B.
(
2012
).
International water conflict and cooperation, the role of power relations among riparians
.
International Journal of Humanities and Social Science
2
(
11
),
56
66
.
Warlick
J.
(
2014
).
Ambassador James Warlick tweet
. Available at: https://twitter.com/AmbJamesWarlick/status/468032136184745985 (accessed 12 November 2015).
Wolf
A. T.
Yoffe
S. B.
Giordano
M.
(
2003
).
International waters: identifying basins at risk
.
Water Policy
5
,
29
60
.
World Health Organization
(
2003
).
The Right to Water, Health and Human Rights Publication Series
.
World Health Organization
,
Geneva
.
World Press Freedom Index
(
2014
).
Yoffe
S. B.
Wolf
A. T.
Giordano
M.
(
2003
).
Conflict and cooperation over international freshwater resources: indicators of basins at risk
.
Journal of the American Water Resources Association
39
,
1109
1126
.